PALEOBOTANY BEERY. 339 



sarcotesta giving the impression the appearance of having narrow 

 wings. This discovery is of special interest since typical Pecopteris 

 fronds have been shown to have constituted the foliage of Psaronius 

 (an undoubted fern) and they frequently show fern fructification of 

 the Asterotheca or Scolecopteris type. A second fern, Aneimites fer- 

 tile, from the Appalachian Pottsville had small bilaterally sjmimet- 

 rical seeds on the somewhat reduced pinnules, the fleshy sarco- 

 testa giving the impressions the appearance of having narrow wings. 

 Other associations of seeds and fronds not found in organic union 

 are the frond species Eremopteris artemisaefolia and the seeds Sam- 

 aropis acuta, the frond species Aneimites bellidulus, and the seeds 

 Lagenospermum Arberi. 



It has been suggested that the Gondwana genera Glossopteris and 

 Gangamopteris were Pteriosperms, but there is as yet no basis for 

 this view. The peculiar Permian frond genus of Asia and North 

 America known as Gigantopteris is found in constant association in 

 the latter country with flat cordiform alate seeds borne in the con- 

 cavity of an obovate reduced pinnule with the Gigantopteris vena- 

 tion, and associated with these are detached two ranked strobiloid 

 organs consisting of similar bract-like pinnules bearing small oval 

 pendent microsporangia — not true cones, but definitely more stro- 

 biloid than any demonstrated Pteridospermophyta. 



It is evident that the Pteridospermophyta constituted a large 

 and varied plexus of synthetic seed plants, our knowledge of which 

 is too imperfect to warrant an extended discussion of their phylo- 

 genetic relations. That they were descended from the ferns is 

 obvious, but whether from the more complex Ccenopteridae (e. g., 

 Asterochlaena), through the polystelic Medullosaceae or from a 

 simple protostelic form of Botryopteraceae through Lyginopteris 

 is debatable. They seem to show an as yet undemonstrated re- 

 lationship with the Paleozoic Marattiaceae. Within the phylum, 

 Scott regards them as showing two lines of descent: One through 

 the evolution of a single stele and leading in the direction of the 

 Cycadophytes, and the other toward polystely which eventually 

 became extinct. Chodat, on the .other hand, regards Lyginopteris 

 as a true fern with specialized megasporangia without higher issue, 

 while he considers the Medullosaceae as the ancestors of the Cyca- 

 dophyta. 



"Whatever the details of descent turn out to be the Pteridosperm- 

 ophyta in some of their forms undoubtedly stand in an ancestral rela- 

 tionship to the Cycadophytes, and the)' may well have merged in 

 their earlier manifestations with the same stock that gave rise to 

 the Cordaitales and Ginkgoales. 



